Innovation Pharmaceuticals Inc (IPIX)

[KMBJN]

One could present risk reduction as absolute or relative.

The risk of developing severe OM without any treatment/placebo is 60(in our study)-75%(upper range of historical), depending on the study and/or population.

With brilacidin, the risk drops to ~40% (37% PP = 7/19 or 43% = 9/21 ITT) - not quite as good as the 22% (2/9) from interim results (as others have noted, the second half of subjects after interim showed approximately equal incidence of SOM with brilacin or placebo, around 50% each).

The absolute risk reduction seems to be about ~30% (range 20-40%), again less than the ~50% (70% placebo 7/10 vs 22% brilacidin 2/9) at interim.

Drug companies almost always present the risk reduction as a relative risk reduction, with bigger numbers. Instead of saying there was a 17.1% (ITT) or 23.2% (PP) absolute risk reduction, they say there was a 28.5% (ITT) or 38.7% (PP) relative risk reduction.

So if the odds are 50% you get some disease, but take a drug and only 40% chance of getting the disease, the absolute risk reduction of 10% is smaller than the 20% relative risk reduction.

This can be a problem when talking about preventing something with a very low prevalence, such as Parkinson's disease, with say a 5% lifetime prevalence. If there was a drug that one could take to cut the risk to 3%, that's only a 2% absolute risk reduction, but a 40% relative risk reduction! Would you rather pay for a drug that they say cuts your risk by 2% or 40%?

For many reasons, some prefer the number needed to treat (which someone also presented here hypothetically) to prevent some disease instead.

Our number needed to treat would be around 3 patients would need to be treated to prevent 1 of them from getting SOM.

I think IPIX was correct in reporting both the absolute and relative risk reductions.

http://onlinelibrary.wiley.com/doi/10.1046/j.1524-4733.2002.55150.x/full

https://www.ncbi.nlm.nih.gov/books/NBK63647/

http://scienceblog.cancerresearchuk.org/2013/03/15/absolute-versus-relative-risk-making-sense-of-media-stories/

https://www.healthnewsreview.org/toolkit/tips-for-understanding-studies/absolute-vs-relative-risk/


Regarding whether the total risk of developing SOM with B can all be called "prevention," would have to agree in principle that we are only "preventing" those who would've gotten SOM without intervention, so should subtract out the placebo response. Our drug "prevented" around 30% of all subjects from getting the disease. Say 30-40% would NOT have gotten SOM anyways, and 30-40% will get SOM even with our treatment.

Still, preventing 30% of 0.5-1.0 million or whatever the annual incidence is for SOM in the developed world, that's still a lot of good outcomes for a LOT of people! Significantly improves their QOL by removing pain, infection, treatment interruptions, feeding tube, hospitalization, etc...


P.S. Sorry for the long didactic post, but got a PM about all the numbers and thought I'd put in my 2 cents. Hope it and the articles help.